[
https://issues.apache.org/jira/browse/AURORA-43?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13892902#comment-13892902
]
Tobias Weingartner commented on AURORA-43:
------------------------------------------
Whoops, forgot about nuking -s and -b. Done.
I'm unsure about the time frame. I think it might be a good thing to have.
Possibly even to require the state to be in deactivated, before you can do a
drain on the host. So if you do:
{noformat}
aurora host_deactivate --duration=3600 cluster somehost
sleep 4800
aurora host_drain cluster somehost
# ERROR, somehost is not in deactivated state
{noformat}
The default duration could be "forever"... Thoughts?
> Rename maintenance commands to have a consistent prefix
> -------------------------------------------------------
>
> Key: AURORA-43
> URL: https://issues.apache.org/jira/browse/AURORA-43
> Project: Aurora
> Issue Type: Task
> Components: Reliability, SRE
> Reporter: Tobias Weingartner
> Priority: Trivial
> Labels: newbie
>
> Let's make the drain/maintenance commands have a common prefix, preferably
> something short. I propose the following in a man page format:
> {noformat}
> SYNOPSIS
> host_status [-f filename] cluster [host ...]
> host_deactivate [-f filename] cluster [host ...]
> host_drain [-f filename] cluster [host ...]
> host_activate [-f filename] cluster [host ...]
> DESCRIPTION
> host_status
> Print the drain status of each supplied host.
> host_deactivate
> The list of hosts is marked for maintenance, and will be de-prioritized
> from consideration for scheduling. Note, they are not removed from
> consideration, and may still schedule tasks if resources are very scarce.
> Usually you would mark a larger set of machines for drain, and then do
> them in batches within the larger set, to help drained tasks not land on
> future hosts that will be drained shortly in subsequent batches.
> host_drain
> The list of hosts is drained and marked in a drained state. This will
> kill
> off any tasks currently running on these hosts, as well as prevent future
> tasks from scheduling on these hosts while they are drained.
> host_activate
> The list of hosts is marked as not in a drained state anymore. This will
> allow normal scheduling to resume on the given list of hosts.
> OPTIONS
> -f, --file filename
> Provide a filename (you can specify stdin using "-") containing the list
> of
> hosts to perform the given action on. No hosts performs nothing.
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
